Cerebral palsy (CP) is the most common class of childhood neuromotor disabilities, resulting in lifelong impairments in productivity and health as well as high costs. To date, no form of infant therapy has proven efficacious via an adequately powered RCT; further, usual and customary therapies often fail to produce clinically meaningful benefits. This revised application builds on new findings from 2 independent preliminary studies showing benefits for infants who received alternative forms of constraint-induced movement therapy (CIMT). Using a multisite RCT design, 72 infants (6 - 18 mos old) with unilateral or asymmetrical CP will be randomly assigned to receive one of 3 manualized forms of multi-component therapies. Primary outcomes include objective assessments (at baseline then 1 wk, 6 mos, and 12 mos post-therapy) of affected upper extremity (UE) skills in unilateral and bilateral activities, as well as changes in brain lateralization based on a novel fNIRS protocol fo infants. The specific aims are: 1) to compare the effects of 3 promising forms of Infant-CIMT that are identical in dosage (3 hr/day X 21 days) and their key therapy elements (shaping, massed practice, home-based, embedded in play and everyday activities, provided by a trained therapist in partnership with parents) but that differ in their use of constraint - continuous casting, part-time splint, or no constraint; and 2) to assess stress levels and safety related to constraint condition (continuous or part-time) compared to no constraint. Stress will be assessed using multiple salivary cortisol samples from infants and parents, parent self-report, and observations. Safety monitoring will include range of motion, skin integrity, sensory awareness, functional use of the casted arm and hand, as well as changes in laterality scores involving the casted side. Even though prior studies affirm the general safety and acceptability of using constraint with infants, there are strongly competing hypotheses about its potential benefits versus risks. Impact: if one or more of the tested infant therapies produces large magnitude benefits lasting up to 12 mos later, then these findings would have a high impact on future clinical practice. Moreover, highly effective therapy very early in life may lead to much higher levels of independence, learning, social engagement, and quality of life for children with unilateral and asymmetrical CP.